Role of Public Finance

Stabilizing Supply
Allocation of Resources
1. Through fiscal and monetary incentives / disincentives government can influence resource allocation
decisions.
2. It can also be used to mobilize resources.
Social Goods
(a) Social Goods and Market Failure
Q. The 'non-rival nature' of social goods consumption has important bearing on efficient resource allocation.
Explore the problem with the examples and diagrams. (2011, I, 20)
1. Market failure due to non exclusiveness: Most social goods non exclusive in nature i.e. there is no
feasible way for A to exclude B from using a social good even when B's consumption affects A's
consumption. Due to non exclusiveness nature of most public good it is generally not feasible to
charge each user. Thus on a busy road, use by B affects use by A but it is impractical to practice
exclusion. So no one will pay for it (since one can enjoy it without paying) and market failure will occur.
2. Market failure due to non rival nature: Social goods are non rival in nature. This means that
consumption by A doesn't affect consumption by B and thus marginal cost of providing social goods is
zero. Thus they can't be provided efficiently (efficient provision means P = MC) by private sector even
when they are excludable and it is possible to charge them (say entry into a public park). The non rival
nature of the social goods mean that the marginal cost of providing that service is zero. So efficiency
principle will dictate that nothing should be charged. But then total costs need to be recovered and in
this case there will be a market failure. The efficient way to do this is to charge a lump sum tax on the
residents for the use of the social good in question.
Consumption Exclusion Exclusion
Feasible Not Feasible
Rival Private Public
Non Rival Public Public
(b) Efficient Provision of Social Goods - Market Mechanism
1. Comparison with private goods: Private goods are rival. Thus if a quantity is consumed by A it can't be
consumed by B. So if we have the MU curves of 2 customers (A and B) for a private good (shown in a)
then to determine the total demand we add up horizontally. Thus we get MUA+B find MC, decide the
price (P = MC) and the quantities consumed by each individual are where the horizontal price line (P)
intersects their individual MU curves. On the other hand social goods are non rival. This means if a
quantity is consumed by A it can be consumed by B as well without prejudicing A's utility in any way.
So if we have MY curves of A and B then this is what they are willing to pay individually for the social
good. Since same quantity can be given to both, we add up their curves vertically to find the MU curve
of the society. By intersection of MC with MUA+B we can get the efficient quantity. Then we see the MU
of each individual at this quantity and charge them that (PA and PB). This is called Lindahl equilibrium.
Note that if output is less than Q then it MC < marginal benefit to society and hence it would be
beneficial for the society to produce more.
2. But here we assume that MU curves of both consumers are known just as they would be known for
private goods.
(c) Issues in Market Mechanism Based Efficient Provisioning of Social Good
1. Free rider problem: If an individual is only one member of a large group and the total available supply
to him is not affected to any noticeable extent by his own contribution then he will have no incentive to
reveal his true marginal utility of the service. Thus he will free ride and will only reveal negligible
marginal utility. The MU curves of A and B will never be revealed. Thus tax on a voluntary basis or as
according to the marginal utility can serve no purpose.
2. Voting: Due to the free rider problem we need a voting based mechanism to reveal society's
preferences as to the necessity of a social good and then to furnish resources for it. If an individual
knows that he will have to comply with the majority decision then he will find it in his interest to vote for
an outcome which is closer to his interests and in this way will reveal his preferences (or MU curve). To
bring out the personal preferences efficiently the voting should be directly linked tax and expenditure
decisions (i.e. it should tell this is the social good we provide, this much quantity and this much is the
tax). In this case voters will reveal their MU curves.
(d) Clarke Groves Tax for Revelation of True Preferences
Voter Option 1
Utility
Option 2
Utility
Option 3
Utility
Tax
Charged
Net Benefit
of Voting
After Tax
A 50 20 10 25 15
B 10 60 20 0 0
C 40 10 55 20 10
Total 100 90 85
Total Without Individuals
Vote
A: B + C 50 70 75
B: A + C 90 30 65
C: A + B 60 80 30
1. If A had not revealed his preference, option 3 would have been selected. By revealing his preference
option 1 is selected but in doing so loss in utility of B + C = 75 - 50 = 25. So A will have to pay a tax of
25. But his gain in utility = 50 - 10 = 40. So even after paying additional tax he is better off by 25.
2. B'd preference revelation doesn't change the outcome so he doesn't pay any tax.
3. C's preference revelation changes the outcome from option 2 to option 1 which entails a loss of utility
by others of 80 - 60 = 20. So he pays a tax of 20. But his gain in doing so is 40 - 10 = 30. So he is still
better off by 10.
Mixed Goods - Externalities
(a) Positive Externalities in Consumption
1. Let MUp represent marginal utility of the good as perceived by its consumers. It will be derived by the
horizontal addition of the MU curves of all individuals. Let MUext be the marginal utility of the good as
perceived by others reflecting the positive externalities of the good. Note that MUext curve will be
obtained by the vertical addition of all the individual MUext curves. We add MUext to MUp vertically to
get MUs i.e. the total MU or benefit to the society.
2. Failure of market: Now a private market mechanism will only equate MUp with MCp and produce Qp of
the good. This is because individuals will be ready to pay only according to the MUp curve. But this is
inefficient from society's point of view because the actual benefit to the society is MUs and hence Qs
would be an efficient production.
3. Need for government intervention: So to expand the output from Qp to Qs government will need to give
either a subsidy to the producers to shift their MC curves lower or give subsidies to the consumers for
this purpose. This subsidy should be equal to MUext. The burden of subsidy is financed by taxes on
individuals - ideally in line with their individual MUext curves but because of the free rider problem - on
the basis of a voting mechanism.
(b) Negative Externalities in Production
1. Here MCext are the external costs to the society of producing a private good. Left to itself, market will
ignore MCext and produce Qp which is obviously inefficient from society's point of view. Hence to move
to the efficient production (Qs) government needs to impose a tax on the producers so as to raise their
costs to MCs or on consumers so as to reduce the demand curve to MUp - tax. This tax should be
equal to MCext. Incidence on each individual can be determined based on the individual MCext curves.
2. The shaded region is the gain to society from the tax. This is excess of costs over benefits for the units
which are eliminated by the tax. It can be proved using congruent triangles a well.
Pigou vs Coase
1. The above argument that the government must interfere in the case of market failure was advanced by
Pigou. However, Coase challenged it. He argued that the presence of externalities doesn't necessarily
mean an interference by government. After all the government intervention will have costs and if these
costs are higher than the social benefits from intervention then government intervention will not
increases social welfare.
1. He argues that in the case with negative externalities of production, let the socially efficient quantity be
Q0 but the production be @ Q1. If the production were to be reduced from Q1 to Q0 the combined loss
of welfare to the producer and the consumers will be the area ACD. But the gain to the society would
be quadrilateral ABCD. Since ABCD > ACD so it should be possible for the society to bribe off the
producers and consumers so affected. But of course whether such a thing happens or not will depend
on the transaction costs.
2. How the bargaining takes place will depend on how the property rights are defined. If in a factory
polluting the river case the residents downstream have the property right over the river then the factory
will have to pay them to pollute the river. This payment will be MSC - MPC for each unit produced
which will raise the cost of the factory and restore the output back to the socially optimum level. If the
property rights were given to the factory then the downstream residents will have to bribe the factory
the area ACD for not polluting the river.
Earmarking
1. Some experts argue that earmarking a certain portion of revenue for a certain expenditure is not a
good budget practice since it imposes constraints. The actual need may be more than or less than the
amount earmarked. But earmarkings have benefits as well.
2. Some taxes may be linked to some expenditures because the tax payments approximate benefits
derived from those services.
3. Earmarking also helps in preference revelation during voting.
Fiscal Decentralization
1. Although social goods are equally available to all concerned, their benefits may be spatially limited.
Thus while national defence may benefit all, street lights may benefit only the people of a locality. Thus
the issue of fiscal federalism arises.
Market Failure due to Increasing Returns to Scale
1. When the industry is subject to IRS, it is unlikely that it will be competitive. Because it will pay for a firm
to eliminate all its competitors and thus to take advantage of the economies of scale. But under a
monopoly the market will not be efficient (P > MC). So if the government takes it over and starts
producing, the market can be efficient. In the above figure, we can see however that P = MC solution
will lead to a loss equal to the shaded area and production is made unviable. What should the
government do in such a case? If instead of making P = MC, it makes P = AC, it can recoup losses but
then the production will not be efficient. Another solution is P = MC along with lump sum taxes which
will ensure no efficiencies. But lump sum taxes are not feasible. Moreover if someone is not using this
service, why should he pay for it.
2. If instead of looking at this PSU in isolation we consider all PSUs then we have a softer constraint: all
PSUs together shouldn't earn losses though any individual PSU can. Also we rule out lump sum taxes
and want to collect user fee only from the users of the services. The solution will not be efficient but our
endeavor is to minimize the excess burden (or the inefficiency).
3. The difference between the user fee charged and MC is nothing but a tax that the government levies
on the commodity to make sure that the PSUs as a group break even. Ramsey rule for optimal taxation
dictates that such a tax on various commodities should be inversely related to their elasticities so that
the demands for each commodity is reduced in equal proportion.
Distributional Effects of Taxation
Endowment Based Criteria
1. Keep what you can earn in the market.
2. Keep what you could earn in a competitive market.
3. Keep labor (earned) income only.
4. Keep what you could earn in a competitive market given equal positions at the start.
Utilitarian Criteria
1. Total welfare is maximized. Benthem argued that if A derives higher utility from an income then it
should be given to A. Sen's handicap is punished here.
2. Average welfare is maximized.
Egalitarian Criteria
1. Welfare is equalized. This leads to opposite of total utility maximization. Sen's handicapped benefits
here.
2. Welfare of the lowest group is maximized. It was propounded by Rawls. It permits higher tax rates so
long as it increases the income of the poor. If carried beyond a certain point it will reduce the total
income and hence reduce income of the poor. Ex ante people will not know where they will end up so
they will vote for an outcome which maximizes the minima. Hence this criteria was considered to be
fair by Rawls.
3. Categorical equity calls for provision in kind.
Mixed Criteria
1. Welfare floor is set with the endowment rule applicable above it.
2. Distribution is adjusted to maximize the welfare in line with social welfare weights.
Limits to Distribution
1. Size of the pie: If we charge higher marginal tax rates then the size of the pie i.e. the total available
income to redistribute may shrink. This is because higher marginal tax rates after a point may reduce
the incentive to work. As we increase the tax on rich they may initially work harder but after certain tax
rate will reduce their work.
2. Efficiency costs: Levying a tax on A may leave him with greater welfare loss than T (tax collected) and
transferring it to B may cause more deadweight loss to the society. However it must be noted that the
fact that the donor loses more than the recipient gains needn't lead to a welfare loss for the society for
it would depend on the social welfare function.
Development
1. We have incremental capital output ratio (v) = (!K/!Y) = (I/!Y). Since GDP growth rate (g) = (!Y/Y),
we have I/Y = v.g. Assuming closed economy, I/Y = S/Y where S is the savings. Now S = Sp + Sg
where Sp is the private savings and Sg is the government savings. Now Sp = mps. (Y - T) = mps.(1 -
t).Y and Sg = (t - a).Y where a is the expenditure of the government as a fraction of GDP. Thus v.g =
(mps - mps.t + t - a) or t = (v.g - mps + a) / (1 - mps) is the required tax rate.
Public Revenue
Efficiency Effects of Taxes
1. The burden of tax is more than the amount of tax collected. Two obvious sources are administrative
costs and compliance costs (resources spent to comply with taxes).
Excess Burden & Efficiency of Lump Sum Taxes
1. In the above case we can see that the imposition of the commodity tax has taken the consumer from a
higher welfare to a lower welfare. While earlier in exchange of OB2 amount of X a consumer was
getting B2G amount of Y now he gets only B2E' amount of Y so that GE' is the tax collected by the
government. The question is was it possible for the government to collect the same amount of tax by
inflicting a lower utility loss on the consumer?
2. If we shift the price ray parallel inwards until the point it just touches the same new indifference curve,
we get to point E". Now @ E" the consumer is at same utility as E' and yet the revenue collected by the
government is E"M which is > GE' (see parallel lines). This means that original taxation was inefficient
and E'N is the excess burden. This arises from the substitution effect. The direct inward shifting
reflects the income effect and is equivalent to a lump sum tax. Thus lump sum taxes don't cause any
excess burden or dead weight loss.
3. But lump sum taxes are not equitable and if we insert an element of equity in it (say a person with
higher income pays more) then it loses its efficiency property.
If the demand for a commodity doesn't change when it is axed, does this mean there is no excess burden?
1. In the above figure we can see clearly this is not the case. Its just that the income effect of the tax has
been completely offset by the substitution effect.
Excess Burden Measurement in a Partial Equilibrium Setting
1. A compensated demand curve is the demand curve which shows the changes in demand only due to
the substitution effect. Thus it is derived after negating the income effects.
1. In the figure above the excess burden is the shaded region. Its area is 0.5*(P.Q.!t^2) / (1/"d + 1/"s).
We know the area of the triangle is 0.5 * base * height. Base = !Q, height = !t.
2. Marginal excess burden can be used to determine the threshold benefit of the project as this
represents the loss to the society by the tax imposed and hence should be the minimum benefit from
the project.
3. But partial equilibrium analysis doesn't give us correct answers. Consider a setting where if a product
is taxed then the consumption of its substitute will increase. That will change the welfare associated
with the substitute market which needs to be taken into account (since it has been caused by
imposition of this tax). In another setting consider a tax on a polluting technology. Taxing it improves
social welfare (since it had negative externalities earlier). But since its prices go up it will create a
distortion in other products as well because of its income effect (real incomes go down).
Excess Burden of a Subsidy

1. In the above figure we can see that the supply curve shifts down when a subsidy is introduced.
Quantity consumed will increase but as we can see it will create an excess burden equivalent to the
shaded area. This could have been eliminated by direct transfer of income. The general equilibrium
graph is the reverse of the "excess burden of a tax" graph.
Excess Burden of Differential Input Tax
1. Lets say a person can work either @ home or @ market. Without any tax both markets will be in
equilibrium @ a where VMPh = VMPm. This is because a worker is paid the value of his marginal
product. Now when a tax t is imposed on the income from market, a person will now equate the post
tax wage rates. Thus we move to e.
2. @ home, initially work done was OKaH. After tax the work done is OKeH'. So additional work done @
home is aeH'H. @ market initially work done is M'MbH'. Post tax work done in market is O'MbH'. So
work lost is abH'H. This means net loss to the society is area abe. It is clear that the higher the
elasticity of VMPm curve and the VMPh curve, higher the excess burden.
Efficiency Effects of Taxation - Choice Between 2 Products
1. In the partial equilibrium figure above the amount of tax collected is in pink shaded area. Grey shaded
area is the dead weight loss.
2. In the general equilibrium setting let AB be the price line under zero taxes and the community is @ E
so that MRSx,y = MRTSx,y and community is @ highest possible indifference curve (IC0). Now if a
direct tax or a lump sum tax or a general consumption tax is imposed the PPF will shift inwards, price
line will shift parallely to A'B' (because head tax will not disturb relative prices). So the community is @
Ed where MRSx,y = MRTSx,y (still). The new indifference curve is ICd which is lower than IC0 because
total available income with society has gone down. EdD is the revenue collected by the government
from product X. If instead of a head tax, a general consumption tax is imposed situation will remain
same.
3. However when a selective tax is imposed, prices will be distorted and society will reach point Eind.
Here MRTSx,y # MRSx,y. Society is @ an even lower indifference curve because distortion of prices
have produced inefficiencies in consumption as well. Thus a general consumption tax is better than a
special consumption tax. @ Eind, the government is now collecting a revenue of EindH. It could have
collected the same revenue EdD by imposing a lump sum tax / direct tax / general consumption tax
and yet left the consumer on a higher indifference curve. Thus excess burden is generated and is
equal to RS.
Efficiency Effects of Taxation - Choice Between Goods and Leisure

1. Imposition of tax has lead to creation of a dead weight loss (grey area) as individuals work less.
2. The general equilibrium setting shows that without tax the society was @ E and indifference curve IC0.
If a lump sum tax is imposed (or somehow it is possible to include leisure in the tax base) then the
community will be @ E' and lower indifference curve IC'. However if an income tax (proportional in this
case) is imposed it is equivalent to imposing a special consumption tax on income where the 2 goods
are income and leisure. Thus the price of income rises and distortions arise and thus community
moves to a lower indifference curve ICt @ Et. In this case the revenue collected by the government is
EtH which is same as E'D. Thus the government could have collected the same revenue by a lump
sum tax and yet leave the individuals on a higher indifference curve. RS is the excess burden
generated by the tax. Thus a lump sum tax is better than an income tax. It can be seen that a
progressive tax rate structure will further distort it.
3. In the case of transfers (which may be treated as a negative case) income effect will be negative.
Substitution effect may be positive or negative depending upon whether it is linked positively to work or
not.
Efficiency Effects of Taxation - Choice Between Consumption and Savings

1. In the general equilibrium setting let the current consumption be on Y axis and the future consumption
(i.e. savings) be on the X axis. Initially let he economy be @ E and the future consumption = (1 + i) *
current consumption. This ensures efficiency and the society is @ E enjoying indifference curve IC0.
Now if a general consumption tax is imposed there will be an inward parallel shift in PPF curve
because a general consumption tax doesn't change the relative prices of savings and current
consumption. So community will be @ Ed running @ indifference curve of ICd. There is no excess
burden. However an income tax will distort the relative prices (as it taxes savings as well as income on
savings) and the community will move @ Eind and even lower indifference curve ICind. In this case
MRTS savings, consumption = (1 + ig) while MRS savings, consumption = (1 + in) where ig is the pre tax rate of
return on savings and in is the post tax return on savings. The government collects a revenue of EindH
in this case but the same could have been obtained via a general consumption tax EdD and still left the
individual on a higher indifference curve. RS is the excess burden of the income tax in this case. Thus
a general consumption tax is better than an income tax.
Efficiency Effects of Taxation - Multiple Choices & Optimum Tax
1. In the above we have only dealt with choice between individual pairs keeping other constant. However
in a general setting, only a direct head tax may be free of excess burden. Also the more elastic a
product's demand, the more is the excess burden on taxing it.
2. Suppose we allow multiple choices like product X, product Y and leisure are all allowed to vary. A
general consumption tax will not create distortions between X and Y but will create distortions between
X and L and Y and L. A special tax on X may reduce distortions between X and L but may increase
those between X and Y. This is specially true of X is a leisure good so that taxing it will make leisure
more expensive and turn people towards more work. Thus in a general setting there is no guarantee
that a consumption tax is a superior system than a special tax.
3. Similarly if savings, current consumption and leisure are allowed to vary it is not necessary any more
that consumption tax is superior than income tax.
4. Consideration for elasticities suggests that higher the elasticity and cross elasticities, lesser should be
the tax. But an optimum tax rule based on minimizing excess burden will suggest higher tax on
inelastic items (like bread) which are used by poor and a lower tax on elastic items (like caviar) which
are used by rich.
Efficiency Effect of Taxation - Tax on Rent
1. Supply of land in the economy is inelastic. So a tax on land rent doesn't lead to any dead weight loss.
Efficiency Effect of Taxation - Choice Between Investments
1. It leads to a transfer of resources from sectors which are taxed to sectors which are not taxed. As can
be seen above this leads to an excess burden equivalent to the shaded area.
Tax Rate, Revenue, Excess Burden
1. From above figures it may be tempting to suggest that a good tax would be the one which maximizes
the distance between revenue and deadweight losses. But social welfare weights should also be
considered.
Efficiency Effect of Taxation: Optimum Taxation - Ramsey Rule
1. Let us assume that the objective of the government behind the tax (on commodities) is to finance the
state's expenditure fully with a minimum of excess burden and without using any lump sum taxes. Also
if it is possible to tax leisure such a tax would have same effect as a lump sum tax. Since it is not
possible we exclude it.
2. Ramsey argued that to minimize overall excess burden, given two unrelated commodities, the marginal
excess burden of the last unit of revenue raised from each commodity must be the same. Otherwise it
would be possible to tax the lower excess burden commodity and release the higher tax burden
commodity. Note that the assumption of unrelatedness is necessary to avoid cross impacts. We also
assume flat supply curves for simplicity.
1. Let initial price be P0 and quantity consumed be X0. Now a tax (t) is imposed such that the new price is
P0 + t and new quantity is X1. In such a case, area 4 represents the excess burden. Now suppose we
raise the tax by 1 unit. The new price will be P0 + t + 1 and new quantity will be X2 such that X2 - X1 =
!x. The new excess burden is area 3 + 4 + 5 such that marginal excess burden is area 3 + 5. Area 5 =
!x.t and area 3 = 0.5 * !x. Thus the marginal excess burden is !x.t + 0.5 * !x.
2. Note that excess burden minimization requires equal excess burden across commodities for the last
unit of revenue collected from them. By imposing additional 1 unit of tax total revenue collected = area
1 + 2. Before the imposition of additional unit tax, total revenue collected was area 2 + 5. This means
imposition of this unit tax has given additional revenue of area 1 - area 5. Area 1 = X2. area 5 = !x.t =
(X1 - X2).t. Thus additional revenue collected (by the unit tax) = X2 - !x.t = (X1 - !x) - !x.t = X1 - !x .
(1+t).
3. Now let !X = X0 - X1. Given the demand curve is a straight line, we have !x/1 = !X/t or !x = !X/t. Also
t >> 1, so we can take 1 + t ~t and thus additional revenue collected by the unit tax = X1 - !x.t = X1 -
!X. Similarly marginal excess burden = !x . (0.5 + t) ~ !x.t = !X.
4. Dividing the marginal excess burden by marginal revenue we get, marginal excess burden of last unit
of revenue collected = !X / (X1-!X). This should be equal to !Y / (Y1 - !Y). This means (!X/X1) =
(!Y/Y1) is the optimum tax condition. But !X/X is nothing but the % change in X. This means Ramsey
rule implies that to minimize the excess burden, tax rates should be set so that the % reduction in the
quantity demanded of each commodity is the same @ the given tax rate on each commodity.
5. Inverse elasticity rule: The price elasticity of demand "d = % change in quantity for unit % change in
tax or % change in quantity = "d * % change in price. And because % change in price is nothing but
the tax rate applied (assuming flat supply curves), % change in quantity = "d * tax rate. Since %
change in quantity has to be the same, this means that items which have higher price elasticity of
demand should face lower tax rates. But this may not be fair from equity point of view.
6. Obviously Ramsey rule doesn't ensure vertical equity. Any departure from Ramsey rule must depend
on 2 considerations. (a) How much the society values equality i.e. how much excess burden is it willing
to tolerate to bring about a given change in distribution or what are the relative weights attached to the
utilities of the poor in the swf? (b) How much the consumption patterns of the rich and the poor differ?
If both rich and poor consume a commodity in same proportion to their income then equity can't be
achieved by differential commodity taxation.
7. If we apply Ramsey rule in an income tax setting we will get that higher the elasticity of supply of labor
the lower should be the income tax rate applied.
The Time Inconsistency Problem
1. Assume that in a society there are only 2 commodities X and Y and that supply of labor is fixed.
Further lets say its © disallows any taxation of Y and so the society taxes only X. Obviously the optimal
taxation rule will dictate some taxation on Y as well and its clear system is inefficient. So we should tax
Y as well and reduce taxes on X.
2. But suppose the citizens are suspicious that after taxing Y the government will not reduce taxes on X.
Rather they will simply use the opportunity to increase their total revenues. So what may appear sub
optimal or irrational behavior on the part of the society may actually be rational. This is called the time
inconsistency problem. In another case suppose the government decides to impose a lump sum tax on
capital and such lump sum tax is only one time. Then it should not alter the savings pattern or the flow
of capital. But people may lack the faith in the government that it will not renege on its promise and
may reimpose such a tax in future as well. So they might alter their behavior.
Supply Side Effects of Fiscal Policy
Effects on Labor Supply - Income Tax

1. Imposition of income tax leads to an 'income effect' which induces labor to work harder (so as to
recoup some lost income). But it also generates 'substitution effect' which works in the opposite
direction (as leisure becomes cheaper). Generally the negative substitution effect outweighs the
positive income effect (since labor supply curve is positive sloping) and higher tax means fall in labor
supply.
2. The general equilibrium setting shows that without tax the society was @ E0 and indifference curve
IC0. If an income tax (proportional in this case) is imposed its equivalent to imposing a special
consumption tax on income where the 2 goods are income and leisure. A person moves to a position
where leisure is higher (assuming substitution effect outweighs income effect).
3. Any person will work less under a progressive tax structure than a proportional rate if the same amount
of tax is to be paid in both cases. Yet work effort of the taxpayers as a group needn't be lower under a
progressive schedule. The net effect depends on how the wage earners at various points on the
income scale respond. Individuals @ the higher wages may be less responsive to changes in tax since
other forms of motivation may dominate. Employees at lower end will be more discouraged but they
may have less flexibility of work hours and they would face lower marginal tax rates.
Effects on Labor Supply - Sales Tax
1. Higher general sales taxes also reduce net income of workers and their work effort will reduce (so long
as they don't suffer from money illusion). But they may also decide to save more and avoid paying the
sales tax. So the question would be whether leisure is traded more readily for present or future
consumption and there is no direct answer.
2. Selective sales taxes will increase leisure if they are imposed on items used in work. If imposed on
items used for leisure they will increase work effort.
Effects on Labor Supply - Transfer Payments
1. They can be regarded as negative taxes. The income effect is now negative and goes on to reduce
work effort. The direction of substitution effect depends on whether transfers rise or fall with income. If
they rise with income (say wage subsidy) then substitution effect will be positive and given an upward
sloping labor supply curve we would expect the work effort to increase. But if the transfers decrease
with rise in wealth (say welfare payments) then the substitution effect will be negative as well and the
overall policy will reduce work effort.
Effects on Labor Supply - Public Services
1. A general neutral public service will have a neutral effect. But like a specific sales tax, if the public
service is leisure promoting then work effort will reduce and if the public service so provided is work
promoting then work effort will increase.
Effect on Investment - Profits Tax
1. As we can see investment falls (because net profit after tax falls) and the higher the elasticity of the
savings and the investment schedules, the higher the fall in investment.
2. In a tax free world, rg = i + d where rg is the gross rate of return required to make an investment, i is the
cost of borrowing and d is the depreciation rate. When we introduce taxes, rg.(1-t) + c = i + d where t is
the tax rate, c is the investment credit. So to increase investments, cost of capital must be reduced.
This can be done by reducing t, increasing c or increasing tax deduction on depreciation. These
different approaches may appear to have same effect but it is not. A change in investment credit will be
limited to new investment while change in tax rate can't be limited. Since tax relief on old investments
has no fresh incentive so government may prefer giving investment credits. Alternatively giving higher
deduction for depreciation encourages long term investments while investment credits favor short term
investments (so that they can be availed off repeatedly).
Effect on Household Savings - Income Tax
1. One might expect that a progressive income tax will reduce the overall savings (because rich have
higher savings rate). But the impact may be less than what one might assume because the difference
in mps of the rich and the poor is < difference in average savings rate of the rich and the poor.
2. Income tax will reduce the return on savings and to this extent may reduce the savings rate but there is
a debate whether or not savings is sensitive to interest rate at all. Even the impact of tax rebates on
savings is doubted.
Canons of a Good Tax System
1. Equality: Adam Smith regarded it as every person should pay according to his ability to pay. On this
basis he argued that a tax should be proportional to the income. But modern economists make a
distinction. On the basis of principle of diminishing marginal utility of money income, they call for a
progressive income tax. There are 2 aspects of equity - (a) horizontal equity i.e. those who are equal
should pay equal, and (b) vertical equity i.e. those who are unequal should pay different.
2. Certainty: Adam Smith said that the quantity, time, manner of payment of the tax payable should be
certain and known clearly beforehand and not arbitrary. If this is left to the discretion of tax authorities it
will weaken the incentive to work and encourage corruption.
3. Convenience.
4. Economy: Collection costs should be minimized and no use spreading out resources thin.
5. Fiscal adequacy.
6. Elasticity of taxation: As income increase due to economic growth the government revenue from taxes
should increase more. Thus progressive taxation is recommended.
7. Diversity.
8. Taxation as an instrument of growth: It should mobilize economic surplus, encourage savings.
9. It should improve income distribution.
10. Promote economic stability: Income taxes promote economic stability automatically.
Concept of Equity
Q. "Subjective approach to taxation leads to least aggressive sacrifice principle". Elucidate. Also give the
limitations of this principle. (2009, I, 20)
Benefits Received Principle
1. Citizens should pay taxes in proportion to the benefits they receive from the services rendered by the
government. Thus it is based on quid pro quo.
2. But its obvious limitations are - (a) Its difficult to measure the extent of benefit derived by an individual
for many services. Most of the public expenditure is on non rival goods. (b) It goes against the very
notion of tax. Tax is not a fee. Its only limited applicability can be when we consider the benefits and
taxes at the level of whole society. But in that case it can only give us total taxes which the society
should pay to the government. Or it can also be applied in the case of toll taxes, special levies on
construction of local roads, sewers etc.
(a) Application of Benefits Received Principle - A General Benefit Tax
1. If the social good is not inferior, generally people with higher income will be willing to pay more for the
same quantity compared to people with lower income. Thus peoples with income of $10,000 may pay
$1 for each unit then people of $20,000 income are expected to pay $2 per unit (here it is assumed
people have identical tastes). But if they only pay $1.50 per unit then it is a regressive taxation
structure.
2. Whether such a structure is regressive / proportional / progressive depends on whether the ratio of
income elasticity of demand to price elasticity of demand ("y/"p) is >/ = / < 1. This is because "y =
(!Q/Q) / (!Y/Y) and "p = (!Q/Q) / (!P/P).
(b) Application of Benefits Received Principle - A Specific Benefits Tax
1. This may be the case in toll taxes, local area charges etc. It can be applied easily where the
commodity provided by the government is a private good. Issuance of licenses, airport fees etc. fall in
this category. This can help ease the pressure on general budget.
(c) Application of Benefits Received Principle - Taxes in Lieu of Charges
1. This is the case where imposition of direct charges is desirable but too costly, so a tax is levied on a
complementary product (e.g. tax on petrol instead of charges on using roads). But its equity aspect can
be doubted. While petrol use depends on the distances driven not each mile driven results in same
costs, nor does it require the same capital outlay to build each mile of road. Thus a person who uses a
particular road may end up subsidizing the use of the other road by another person.
Ability to Pay Principle
1. It asks people to pay taxes according to their ability to pay. It essentially means that the 'burden' of tax
falling on everybody should be the same. But how do we objectively measure the ability to pay or the
burden? Should it be the income, consumption, wealth? Income has been the most widely accepted
measure of the ability to pay but theoretical framework yields more (although not unqualified) support
to the consumption as a measure (overall personal consumption and not just a specific commodity
consumption). But whatever base be used it has to be comprehensively defined (no leakages).
2. Progressiveness of a tax can be defined either in terms of marginal progressiveness or absolute
progressiveness. Generally it is taken on the average parameter i.e. the average tax rate should
increase with increase in incomes. Then to measure the degree of progressiveness, there are 2 criteria
- (a) V1 = (t1-t0) / (I1-I0) where t is the average tax rate for a given income and this simply means that
greater the increase in average tax rates as income increases, higher the progressiveness of the tax
structure. (b) V2 = (!T/T) / (!I/I) or simply the income elasticity of the collected tax revenues. It says
that higher the income elasticity of the collected tax revenues, higher the progressiveness of the tax
structure.
Ability to Pay - Vertical Equity & Subjective Approach
1. In this the concept of sacrifice undergone by a person in paying a tax occupies a crucial place. In
paying a tax a person suffers from some disutility. This disutility is the sacrifice made by him. Thus this
approach measures the ability to pay in terms of the loss of utility by the tax payers.
(a) The Principle of Equal Absolute Sacrifice
1. If U is the total utility function, Y is the pre-tax income and T is the tax paid by an individual then U(Y) -
U(Y-T) should be same for all individuals because this is the sacrifice made.
2. If this principle is applied no one will be exempted.
3. Constant MUy: Further if the marginal utility of income (MUy) is constant and not diminishing, then it
would also mean everybody pays same amount of tax.
4. Diminishing MUy: Let the tax be very small so that the shaded areas can be considered as rectangles.
Now loss of utility by poor = MUpoor * Tpoor and loss of utility by rich = MUrich * Trich. Both have to be
equal under this principle, thus MUpoor * Tpoor = MUrich * Trich. In other words, MUpoor/MUrich =
Trich/Tpoor. Now this is an identity under this rule. If tax rate is proportional i.e. T = c.Y where c is a
constant, then MUpoor/MUrich = Yrich/Ypoor i.e. the fall in marginal utility with income is in same
proportion as rise in income. If MUpoor/MUrich is greater than the ratio of the incomes then it will mean
Trich/Tpoor> Y rich/Ypoor or Trich/Yrich> T poor/Ypoor or tax system is progressive. Thus tax system is
progressive if the fall in MU is higher compared to rise in income. Similarly if the fall is less i.e.
MUpoor/MUrich< ratio of their incomes then it means T rich/Tpoor < Yrich/Ypoor or Trich/Yrich < Tpoor/Ypoor or
tax system is regressive.
1. Income elasticity of MUy: Let income elasticity of MUy be " i.e. " shows the % change in MU for one %
change in income. Then the tax system is regressive, proportional or progressive depending upon
whether " is less than, equal to or greater than 1. Proof: Burden of tax paid by poor (in orange) = MU.t
and burden of tax paid by rich (in black) = (MU - !MU) . (t + !t) or MU.t - t.!MU + MU.!t. Both are
equal under this principle. So MU.t = MU.t -t.!MU + MU.!t or t.!MU = MU.!t or (!MU/MU) = (!t/t) and
this is an identity under this system. Thus " = (!MU/MU)/(!Y/Y) can be rewritten as (!t/t)/(!Y/Y). Now
a regressive tax system is defined as one where % change in tax < % change in income i.e.
(!t/t)/(!Y/Y) < 1. But this is nothing but ". So if " < 1 the taxation system is regressive. Similarly cases
of " = 1 and " > 1 can be proved.
(b) The Principle of Equal Proportional Sacrifice
1. This principle requires that every person should be made to pay a tax such that the % loss in utility is
same for all individuals i.e. {U(Y) - U(Y-T)} / U(Y) is constant.
2. Constant MUy: A tax system following this principle will be proportional. Proof: The loss of sacrifice of
poor = area under orange rectangle = Tpoor * MU. Total utility of poor = Ypoor * MU. Thus their ratio =
Tpoor/Ypoor. Since this ratio has to remain same => tax system will be proportional.
3. Diminishing MUy: A tax system following this principle will be progressive.
(c) The Principle of Equal Marginal Sacrifice (Minimum Aggregate Sacrifice)
1. It means that MUy of each person after paying the tax should be same. This approach seeks to
minimize the aggregate sacrifice of the society as a whole i.e. when all persons pay tax such that the
marginal utility of income left after paying the tax is same, the overall sacrifice of the society is
minimum.
2. Assuming MUy curve remains the same it means very high marginal rates of taxes for rich and in fact
equal post tax income for all.
Ability to Pay - Objective Approach
1. It considers the question - What should be the objective base of taxation for measuring ability to pay?
2. Income: Income is generally considered to be the best measure. But the ability to pay increases more
than proportionately to the amount of income (hence the justification of progressive taxation). Further
distinction should be made between the earned income and unearned income and consideration
should be given for the number of dependents.
3. Wealth: It is considered to be a better base since it represents the accumulated purchasing power.
Ownership of assets may give a purchasing power quite different from what money income may
suggest.
4. Consumption: Advocated by Kaldor, it argues that consumption is the amount of resources that a
person actually withdraws from the economy for his personal use. Savings lead to increase in capital
stock and hence increase in productive capacity of the economy. Hence higher the consumption,
higher should be the tax. He thinks this is particularly relevant for developing countries so that they can
promote savings.
Horizontal Equity - Equals Should be Taxed Equally
(a) Which is a better base?
Income
Tax
Income
Tax
Consumption
Tax
Consumption
Tax
Tax on
Wage
Income
Tax on
Wage
Income
A B A B A B
Period 1
Wage
Income
100 100 100 100 100 100
Consumption 90 - 90 - 90 -
Tax 10 10 10 - 10 10
Savings - 90 - 100 - 90
Period 2
Interest - 9 - 10 - 9
Tax - 0.9 - 11 - -
Consumption - 98.1 - 99 - 99
Savings - - - - - -
Total Tax 10 10.9 10 11 10 10
PV (Total
Tax)
10 10.82 10 10 10 10
1. It can be seen that consumption tax gives better horizontal equity than income tax.
2. But consumption tax can be postponed indefinitely. Also it is assumed that people with same present
value of all future income are in equal position but this is difficult to estimate. Also this proposition of
equal PV disregards when the income occurs. For this to happen we need to have perfect capital
markets. One big drawback of consumption tax is that the more successful it is (in reducing
consumption and increasing savings) the higher the tax rates needed to maintain same amount of
revenues.
(b) Comprehensiveness of the Base
1. It is generally agreed that the bases should be comprehensive. This means that income tax should be
levied on all accrued income irrespective of the source, time of realization and consumption tax should
be levied on all consumption. A person's ability to pay is increased whether the income comes in the
form of money income, or capital gains.
(c) Treatment of Bequests
1. Tax payment under a consumption tax can be postponed indefinitely (by postponing consumption) and
savings can be passed on to the next generation. So a tax should be there on bequests. Leaving
savings for the next generation is like using them.
(d) Wealth as Tax Base
1. Wealth can be seen as capitalized value of capital income, thus if capital yields a return of 10% a 10%
tax on the return would mean a wealth tax of 1%. Wealth tax is thus a tax on capital income whereas a
consumption tax would in effect exclude capital income from the income tax. Thus consumption tax
runs against an wealth tax (when paid on top of income tax).
2. But savings and thus wealth generates additional utility. This gain should be taxed as well in addition to
the consumption tax.
3. Income is not the sole judgement criteria for equity. Wealth also gives purchasing power.
(e) Land as a Tax Base
1. The question which arises is what should we tax on land? Should we tax actual income or potential
income or value of land? In perfect markets this question would be irrelevant because value of land will
reflect PV of its potential income and potential income would be equal to actual income. But in real
market this differs. In the below table we can argue that a tax on the potential income of the land is
better than tax on the actual income from the objective of increasing land utilization.
100% utilization 50% utilization 0% utilization
1. Actual Income 100 50 0
2. Cost of underutilization to the owner 0 50 100
Case 1: 10% tax on actual income
3. Tax 10 5 0
4. Post Tax Income 90 45 0
5. Cost of underutilization to the owner 0 45 90
Case 2: 10% tax on potential income
6. Tax 10 10 10
7. Post Tax Income 90 40 -10
8. Cost of underutilization to the owner 0 50 100
1. Thus we can see that in case 1 (when we tax actual income) the cost of underutilization to the owner is
less than the corresponding costs without any tax. So such a tax discourages land utilization. On the
other hand a tax on the potential income (case 2) leaves the cost of underutilization unchanged and
thus is better.
(f) Feldstein's Utility Criteria or the Utility Definition of Horizontal Equity
1. Feldstein gave 2 criteria which a tax must satisfy to ensure horizontal equity. (a) People with same pre-
tax utilities should end up with same post tax utilities. (b) Taxes should not alter utility ordering i.e. if
someone is better off than the other pre tax then he should be better off post tax as well.
2. It can be seen that income taxation doesn't satisfy Feldstein's criteria. Assume a person A who derives
utility out of material consumption say food and a person B who derives utility out of spiritual pursuits.
Lets say before the tax they have same utility levels and same incomes. After the income tax, the
income and hence utility of A will be severely curtailed while in B's case only the income will go down
but the reduction in utility will be lesser. Thus condition 1 is violated.
3. It can also be seen that any pre-existing tax will satisfy the utility definition of horizontal equity if
individuals are free to chose their activities and expenditures. And further that it is only the new taxes
which will lead to a violation of the horizontal equity criteria. Lets say there are 2 jobs - one with lots of
perks (which don't fall under income tax net) and one without any perks (which fall under income tax).
Lets also assume people have same utility functions. Before the tax is imposed the final utilities of both
jobs will be equal (due to migration) so that the wage in the no perks job will be higher than the wage in
the with perks job. Now if we impose any income tax that alone will lead to a change in equilibrium and
violation of horizontal equity. This is because an income tax will now make people in no perks job
worse off than people in with perks job. Thus, "the only good tax is an old tax".
(g) Normative Analysis of Tax Evasion
1. The usual model is that if (the penalty upon being caught * probability of getting caught) < (gains from
tax evasion), then a person will evade tax. But it ignores the psychic costs (guilt feeling, conscience
etc.), ignores the different natures of jobs (in some works it may be easier to evade tax, in some it may
not be) and ignores the risk aversion (risk averse people are less likely to avoid tax).
2. Suppose now we have to formulate a tax policy. Generally it is assumed existence of a black economy
lowers social welfare. Let us construct a social welfare function. An important consideration in
constructing it would be if we should include the welfare of tax evaders as well in our swf? If we do,
then under some conditions, the existence of the black economy may actually increase swf. For
instance according to Ramsey rule we should tax commodities with higher elasticity at a lower tax rate.
But say the government taxes gold (assumed to be highly elastic here) at a high rate. In black
economy its effective taxation will be very low. Thus black economy is reducing the excess burden of
the tax. On the other hand if we chose not to include the welfare of tax evaders in a swf then we will
gravitate towards a rule which imposes severe penalties (limit is a death penalty) on tax evaders.
Obviously in real world a middle path is followed.
Sources
Direct Taxes
(a) Merits
1. Equity: They can be structured so as to closely relate to the ability to pay principle.
2. Distribution effects: They can reduce inequalities.
3. Allocation effects: They minimize allocation distortions. Merely reduce the Yd.
4. Stabilization effects: Act as automatic stabilizers.
5. Elasticity: High income elasticity specially if progressive.
(b) Demerits
1. Reduce incentives to work, save and invest.
Indirect Taxes
(a) Merits
1. Ease in collection, huge revenue potential.
2. Can be used to influence pattern of production and demand.
(b) Demerits
1. Regressive, inflationary.
1. Resource allocation: Create inefficiencies in resource allocations by distorting relative prices and thus
create excess burden. As can be seen imposing an indirect tax on commodity X distorts the price line
for consumers and makes it steeper (PLind vs PL0). Thus consumption will take place @ K where PLind
is tangent to ICindirect. Consumer loses because he is no longer @ IC0 and production is inefficient
because PLind is not a tangent to PPFind. Had a direct tax only been imposed, consumer would have
been @ ICdirect which is better off and production would have been efficient as well.
Incidence of Taxes
Differential Incidence vs Absolute Incidence
1. Tax revenues are generally not earmarked. Thus in our analysis on incidence we will refrain from how
the money so collected is used. While doing the analysis we keep in mind both the source side and the
use side incidence effects of the tax.
2. In differential tax incidence we examine how the incidence differs when one tax is replaced by another,
holding the total revenues (and of course total expenditures) constant. The other tax here is usually the
lump sum tax. In absolute tax incidence we examine the effects of the tax when there is no change in
either other taxes or the expenditure. There is another analysis called budget incidence where we also
analyze the impact of the spending pattern of this tax proceeds.
Incidence Effects of a Unit Commodity Tax - Perfect Competition
(a) General Case

1. Burden shared by consumer = Pgross - P0. Burden shared by producer = P0 - Pnet. Ratio of the burden
shared = |Pgross - P0| / |P0 - Pnet| which is same as (!Pd/!Ps). If we divide both numerator and
denominator by !Q and then multiply both numerator and denominator by (Q/P), we will get ("s/"d).
Thus the incidence tends to fall on the consumer if the demand is inelastic while the supply is elastic
and it tends to fall on the producer when the demand is more elastic than the supply. This also means
that taxes on a group of close substitutes will have more incidence on the consumer since the price
elasticity of demand across a group of substitute is < the price elasticity of demand for a particular
product only (while there is not much difference on the supply elasticities in short run at least). Similarly
elasticity of demand will be higher as longer time period is involved. Similar considerations apply on
the supply side.
2. Thus we see that the burden borne by the households is from the uses side. So such a tax will be
progressive if the income elasticity of demand > 1. Taxes on luxuries are thus progressive in nature
while those on necessities are regressive. A general sales tax on all products will be regressive
because consumption as a percent of income falls while moving up the income scale.
3. To the extent the burden falls on the seller, factor earnings involved in the production of the commodity
are affected. Thus if the commodity uses highly skilled labor or capital then such taxation will be
progressive while if it involves unskilled labor, it would be regressive.
4. In a general equilibrium analysis framework, the substitution effect of the tax will increase the demand
for the other commodity (Y). Now if say X was labor intensive and Y was capital intensive it would lead
to a fall in (w/r). Similarly on the uses side if the production of Y is subject to increasing costs, it would
increase Y's price as well and the uses side impact will be spread over the consumers of Y as well.
5. We can also look at the right panel where instead of bumping up the supply curve, the demand curve
is lowered.
(b) Horizontal Supply Curve or "s tends to infinity
1. Thus we can see that in such a case entire burden will be passed on to the consumer.
(c) Horizontal Demand Curve or "d tends to infinity
1. Here it can be easily seen that the entire burden falls on the producer.
Incidence Effects of a Commodity Tax - Oligopolistic Industry
1. Like the perfect competition and monopoly, firms contract their output. But it may so happen that they
were overproducing before the tax. Now they may reach closer to a cartel solution and hence be
actually better off in some cases.
Incidence Effects of a Payroll Tax
1. Same as the discussion above. Similarly supply and demand elasticities will be larger if tax is selective
rather than general (like a personal income tax).
2. On the uses side if the taxed factor is used more intensively in commodities which are necessities
(income elasticity of demand < 1) then such a tax will be regressive otherwise it may be proportional or
progressive.
3. In a general equilibrium framework, as the supply of the taxed factor (L) reduces the relative supply of
the other factor (K) will increase and in this case (w/r) will increase and thus the burden is shared by
the capital as well. This will lead to a switch over to more capital intensive (higher K/L) techniques. On
the uses side as the prices of commodity X which uses the taxed factor (L) more intensively, its
demand decreases (substitution effect) and relative demand for Y increases whose price will rise as
well thus sharing the uses side burden on the consumers of Y as well.
1. In imperfect markets, question arises will not the unions be able to shift an increase in the income tax
by demanding a higher pay and will not the employer be able to pass on the cost to the consumer in
form of higher prices? As can be seen in the above diagram, this is certainly possible and will depend
on the relative elasticities.
Incidence Effects of a Capital Gains Tax
1. Same as above. It may be noted here that in an economy with effective restrictions on capital flows
(and no black economy), total supply of capital will be inelastic. Hence there will be no escape from the
capital gains tax and the incidence will be on the suppliers of capital (to the extent determined by the
price elasticities). But in an open economy where capital is perfectly mobile, the suppliers of capital will
bear no burden of the tax whatsoever. This is because supply curve of capital will be perfectly flat and
capital has the option to flee to other countries.
2. But even though adjustments can't be made in the SR, in the LR the tax will have impact on behavior
as the outworn capital may not be replaced and previously planned expansions may be canceled etc.
3. On the uses side if the taxed factor is used more intensively in commodities which are necessities
(income elasticity of demand < 1) then such a tax will be regressive otherwise it may be proportional or
progressive.
Incidence Effects of a Profits Tax
1. A profit tax doesn't change MC or MR curves. Thus no firm has any incentive to change its output
decisions. So all tax incidence is on the producers only. Price, quantity, wages etc. remain same.
Alternatively, it can be visualized that while the firms were earlier maximizing profit P, now they are
going to maximize profit P * (1-t). Because they impose no inefficiencies, profits taxes may appear very
attractive. But an obvious problem here is that for such a tax to be efficient it must be on economic
profit and not financial profit. And there are obvious problems in measuring economic profits.
2. When a profits tax is imposed, surely the capitalized value of the enterprise falls (since it is PV of all
future profits) and the person who is the current equity holder bears all the loss. But thats about it.
Future owners are not penalized because even though they will be paying the profits tax in future, they
got the equity at cheaper rate.
3. In the case of a monopoly, the monopolist has no option but to absorb the entire tax. Again this is
because a profits tax doesn't change MR or MC. Earlier while he was maximizing TR - TQ, now he will
maximize (1 - t) * (TR - TQ) and hence no change. He will continue to produce @ same level and price
until he earns profit else he shuts down the shop.
Incidence Effects of a Land Tax
1. Supply of land is inelastic. So the entire burden of taxation will fall on the suppliers and their rent would
be reduced.
Incidence Effects of Corporate Tax
1. Since in the LR the flow of capital between the corporate and the non corporate sector is highly elastic
and the tax is applied on corporate sector only, so the capital will move from corporate sector to the
unincorporated sector. The final net post tax returns in both sectors will be the same. The output in
unincorporated sector would have increased and that in corporate sector would decrease.
2. When the tax is imposed, initially the owners of the capital in corporate sector suffer. But then capital
migrates, return on capital is lowered in both the sectors (depending upon the relative elasticities of the
two sectors) and thus owners of capital in both sectors suffer. It can be seen in the figure above that
the burden (!t) is shared both between the corporate and the non corporate sector.
3. If we assume that the corporate sector was more capital intensive and unincorporated sector was more
labor intensive then the fall in output of the corporate sector releases more capital and less labor. This
increases (w/r) in general and thus labor benefits and capitalists suffer. Moreover there is a general
increase in (K/L) ratio as K becomes cheaper.
4. On the uses side the consumers of products of corporate sector would be burdened while that in the
unincorporated sector will benefit. But if the production in unincorporated sector is subject to increasing
costs then both consumers will share the burden.
Equivalence Between Taxes
1. Let T be a tax on capital used in production of commodity X, TK,Y be tax on capital used in production
of commodity Y, TL,X be tax on labor used in production of commodity X, TL,Y be the tax on labor used
in production of commodity Y, TX be a commodity tax on X, TY be a commodity tax on Y, TK be a
general capital gains tax, TL be a general labor tax, T be the general income tax.
2. Thus it can be seen that TX and TY at same rates are equivalent to a general income tax T. Both
create a parallel shift inwards. Similarly if TK and TL are imposed at uniform rates, it is similar to the
general income tax T as well.
3. Similarly a tax on both labor and capital employed in commodity X @ same rate is equivalent to a
commodity tax on X itself.
4. In an economy without savings a general income tax will have the same effect as a general
consumption tax.
5. In perfect competition it makes no difference on whom the tax is imposed - both in goods and factor
markets.
The Harberger Model
1. His assumptions are as follows - (a) CRS prevails. (b) Perfect competition in all factor and commodity
markets. (c) Perfect mobility of the factors. (d) Total amount of each factor is fixed and is fully
employed. (e) All consumers have identical tastes. (f) X is labor intensive, Y is capital intensive. (g)
And we consider only differential tax incidence here.
2. Incidence of a specific commodity tax (TX): When a specific commodity tax on say X is imposed, its
relative price will increase. Consequently less of X and more of Y will be produced (as demand for X
falls). As this happens factors migrate from X to Y. Now because X was labor intensive, a fall in
production of X releases more L than Y can absorb at the given K/L ratio. So (w/r) falls and both X and
Y become more capital intensive (K/L increases). Thus in the new equilibrium, all labor is worse off
(and not just that earlier employed in sector X) and all capital is better off. Now greater the price
elasticity of demand of X, the greater the fall in production of X and hence greater the decline in
wages. Similarly the greater the difference in (K/L) ratios for the two sectors, the greater the decline in
wages. Finally, the lower the substitutability of labor for capital in Y, the greater the decline in w
needed to achieve the desired change in (K/L) to absorb additional labor. On the uses side people who
spend relatively larger proportion of their incomes on X will be worse off.
3. Incidence of an income tax (T): Since factor supplies are completely fixed, this tax can't be shifted. It
has to be borne by people in proportion of their initial incomes.
4. Incidence of a partial factor tax (TK,Y): Lets say the capital used in production of Y is taxed. This will
have an output effect i.e. the price of Y will tend to increase which will decrease its quantity demanded,
and a factor substitution effect i.e. as capital becomes more expensive so (K/L) decreases. Due to the
output effect, both K and L are released (more of K because Y is a capital intensive industry). So (w/r)
will go up and (K/L) will tend to increase. If we consider the substitution effect then (w/r) will increase
as demand for capital will be reduced. If the factors were immobile between commodities then in this
case capital owners employed in Y suffer because the supply of the capital is inelastic. But it will also
reduce the VMPL,Y and hence overall wages will also suffer while the capital owners in X will benefit
(as their capital now has more labor to work with and hence higher productivity).
Q. How does the burden of tax distribution between buyers and sellers in the ratio of elasticity of demand and
that of supply take place? (2011, I, 20)
Q. The direct money burden of the tax imposed on any object is divided between the buyers and the sellers in
the proportion of the elasticity of supply of the object taxed to the elasticity of demand for it. Discuss. (2009, I,
20)
Q. "The adjustments to a tax imposition not only affect the distribution of tax burden but also bear upon the
efficiency of resource use in private sector."
Substantiate the statement highlighting the role of taxation policy in improving the allocative efficiency in an
economy. (2007, I, 60)
Limits to Taxes and Borrowings
Crowding Out Effects
1. Increase in G leads to rise in Y which increases the transactions demand for money. This reduces the
available money for speculative purposes and thus r rises. Once r rises, I will fall due to a move up on
the Id curve.
Q. If public expenditure is financed by money creation, show diagrammatically the short run and long run
crowding out effect. (2010, I, 20)
Deficit Financing
1. Keynes: The nature of unemployment prevailing in developing countries was different from that
prevailing during recessions in developed countries. In developing countries it was of the type of
disguised, chronic and lack of wage goods type. So raising AD via deficit financing will not be effective
in developing countries. Deficit financing can only help when there are idle resources in the economy.
But the development in last half a century has made AD an important factor in determination of Y and
employment. It can no longer be assumed that the country will not have deficiency of demand.
2. Lewis: He agrees that when the modern sector expands and workers earn higher wages there will be a
rise in prices as the workers are paid out of the new money created. But when the formed capital is put
to use output of consumer goods will also increase leading to a stabilization in prices. Further with the
expansion of modern sector, not only do output and employment increase but also profits. As the share
of profits increase the amount of investment being financed out of the created money diminishes and
ultimately the increase in voluntary savings kills inflation.
3. Deficit financing induces forced savings as due to rise in price levels purchasing power of people
decreases and they are forced to consume less.
4. Structuralist theory (Myrdal): Deficit financing is done in developing countries to finance government
and private sector investment needs. An important bottleneck is lack of resources for financing
economic development. In developing countries governments play a vital role in planning and
industrialization. This requires large resources but the structure of the economy is such that it is often
not possible for the government to raise revenue via taxation or borrowings (lack of administrative
structure, improper institutions, corruption, shallow financial markets, lack of surplus or purchasing
power with people). So governments have to resort to deficit financing. Thus growth in Ms is only the
apparent and not the real cause of $.
5. Non - inflationary deficit financing (Fischer, Easterly): The amount of revenue government can expect
to obtain from printing money without stroking inflation is determined by the ratio of reserve money (H)
to GDP, real growth rate of economy and income elasticity of demand of real balances. Thus if reserve
money is 20% of GDP, GDP growth rate of 10% and income elasticity of demand for real balances is
1.5, then government can raise 10% * 1.5 * 20% = 3% of GDP as revenue without stroking inflation.
Thus in India where growth rate of GDP is high and demand for money is increasing (as economy gets
more monetized) a reasonable amount of printed money can be created.
Effects of Debt Financing
1. Inflation tax revenue: When $ = 0 then obviously $-tax revenues are zero. As $ increases $-tax
revenue also increase. But people begin to hold less and less of real cash balances with them and
more in $-hedged assets as $ increases. Finally at a point the holding of real cash balances with
people declines to such an extent that there is a decline in overall $-tax revenue.
1. Keynesian framework: @ a fixed price level, increase in G will cause increase in AD and thus outward
shift in IS curve. If the economy is @ less than full employment then an output gap exists and output
will also rise in response to rise in AD. As a result for a given tax rate, tax collections will also rise and
ultimately @ given tax rate tax receipts will balance out the !G as economy reaches Y*. But critics
argues that it underestimates crowding out effect.
1. Wealth effect (Patinkin & Friedman): By financing deficit via sale of bonds, government increases the
wealth of the people. This is because bonds are held as wealth by people. As a result of increased
wealth, people will demand more money. This leads to leftward shift in LM curve which offsets any
expansionary effect of increased G. This is the wealth effect but empirical studies suggest its not as
significant.
2. Ricardian equivalence: Consumers look at not their current income but their permanent income or
expected future income while making consumption decisions. This is based on Modigliani's life cycle
theory of consumption and Friedman's permanent income hypothesis. Now if the government
increases deficit and finances it via borrowings, a consumer will know that there is no way to finance it
other than rise in future taxes. This will dampen his future income and thus will not induce him to
increase his present consumption much. They will only increase the consumption if PV (future
additional taxes) < current increase in deficit.
Central Bank's Dilemma
1. If the central bank doesn't monetize government's deficit then government will have to borrow from the
market and hence a rise in interest rates and crowding out effect.
2. If the central bank monetizes the deficit it leads to creation of H and hence inflation. If the economy is
in recession then deficit financing will not be inflationary and in fact raise output.
Q. Bring out the differences in the definition of deficit financing given by different authorities in India and
examine the role of deficit financing as an instrument of monetary control. (2009, I, 20)
Disposing Budget Surplus - Effect on Inflation
1. Using surplus to retire public debt: This returns the money back to public and hence weakens the anti-
inflationary impact.
2. Impounding surplus funds: This will extinguish the money and hence reduce Ms and hence will be
strongly anti-inflationary.
Public Expenditure
Decision Rules
How to Allocate - Divisible Projects
1. If the budget size is fixed, it is easy to see that the optimum utilization of funds (highest possible total
benefits) would be when the marginal benefits from the last dollar spent on each projects is the same
(otherwise it will be possible to reallocate such that the project with higher marginal benefit gets more
at the cost of project with lower marginal benefit).
2. If the budget size itself is variable (i.e. we also have to decide how much to tax) then not only should
the marginal benefits of the two projects be equal but should also be equal to the marginal loss to the
society as the last dollar is taken away from the private sector.
How to Allocate - Lumpy Projects & Fixed Budgets
1. We have different projects with different benefits and different costs. Let rule 1 be we arrange the
projects in terms of their benefits-to-cost ratio in a descending order and then we select projects from
the top until we are within the budget limit. Let rule 2 be we select a mix of projects which gives the
highest net benefit. Let rule 3 be that we try to minimize the left over (residual) subject to the constraint
that projects selected have a cost - benefit ratio > 1.
2. It is clear that 1 and 2 are superior to 3 since both buy more benefits at a smaller cost. Rule 2 can be
preferred over rule 1 if the marginal value of the residual dollars is zero otherwise rule 1 is the best.
How to Allocate - Lumpy Projects & Variable Budgets
1. Here the problem is again to weigh the marginal benefits of the public sector projects against the cost
of the marginal benefits forgone in the private sector.
Burden of Public Debt
Refunding vs Debt Repayment
1. As the debt grows larger and larger, how will it ever be possible to repay it? This question is misplaced
because household debt must be repaid sooner or later. Public debt needn't be repaid, it can be
refinanced since the budget and the economy are a continuing undertaking.
Tax Burden of Debt Service
1. If the debt is held domestically then it can be argued that it doesn't create any excess burden because
even if we impose taxes to repay the debt, the money will remain in the economy itself. This argument
is wrong because the taxes imposed will have an excess burden / dead weight loss on the society
(even if the money stays in the economy).
2. Continuing expansion od the debt combined with a constant GDP will lead to an infinite debt to GDP
ratio. But if there is a constant fiscal deficit to GDP ratio and the GDP is also growing at a constant rate
then the debt to GDP ratio will approach a constant in the steady state.
Intergenerational Equity Aspects of Debt Finance - Reduced Capital Formation in Private Sector
1. In the classical economics, resources withdrawn from the private sector (via debt borrowings by the
government) leave the private sector with fewer resources and thus the debt burden should be borne
by today's generation. But the question is where those resources are coming from? If coming from the
consumption then the present generation's welfare is reduced. If coming from the savings, then the
future generation will have lesser consumption.
2. If we assume that the debt financing is used to create capital assets (which will benefit the future
generations) then the benefits principle argues that it be financed by future generation only. Thus if we
further assume that all borrowings are paid out of savings and all taxes are paid out of consumption
then we should divide the budget into 2 parts - current and capital and finance the current out of taxes
and capital out of borrowings.
3. But the rational expectations approach questions whether individuals will behave differently to taxes
and borrowings. As rational agents they will understand that a borrowing now means heavier taxes in
future. Thus their net worth is reduced (as they will have to pay higher taxes in future). Even those who
have lent to the government have their net worth reduced because of the higher taxes. Thus the
current consumption will be reduced nevertheless.
Intergenerational Equity Aspects of Debt Finance - Foreign Debt
1. Borrowing from foreigners implies that there is no need for the current generation to lower its
consumption. Nor is there a need to lower the investment in the capital stock for the higher
consumption of the future generations. So now the future generations will have unchanged
consumption capacity, but they will have to pay a part of their GDP to the foreigners.
2. The deciding rule should be where the cost of debt is lower. If it is lower for the foreign debt then one
should borrow from outside.
Term Structure and Debt Management
1. If the yield curve is upward sloping it doesn't mean that one should borrow in the near term only. Or if it
is inverted it doesn't mean that one must borrow long term. What matters it the expectations of rate
change. If the treasury expects rates to go up then it must borrow long and if it expects rates to go
down then it must borrow short.
2. But treasury has control over interest rates as well and it can very well print money to service the debt.
But such an action will reduce its credibility and give a shock to the economy. The treasury can always
get resources from the private sector by handing them money. But the purpose of issuing debt rather
than money (or replacing maturing debt with new debt rather than monetizing it) is to purchase
illiquidity. Investors must be convinced to hold debt rather than money and the only way to do this is to
pay them. Issuing LT debt reduces liquidity. So even if the LT rates are higher, it might pay the
treasury to issue LT debt.
3. Lengthening the debt at the time of refunding tends to be restrictive while shortening it tends to be
expansionary. Lengthening raises LT rates which are used in most investment decisions. So this
restrictive effect of LT financing should also be kept in mind.
Excess Burden of Debt: Tax Exemption vs Direct Interest Subsidy
1. Tax exemption can't be justified on vertical equity grounds since the higher income slab people get
higher benefits. Nor can it be justified on the horizontal equity grounds since people with equal
capacity may end up paying different tax depending upon their behavior (one who invests in the tax
free bond pays less). It also leads to a substitution of tax saving for other savings. Thus a general
direct interest subsidy is better.
Effects
Effects on Work Incentives - Welfare Plans
1. In the left panel, we show plan (i) where we give a lump sum subsidy to all people below the poverty
line (PL) and zero to all APLs. The blue line is the 45º line where there are no taxes and subsidies. In
the left bottom panel, equivalent marginal tax rate is shown. It is clear that as soon a dollar is earned
@ PL, the income received falls drastically and this means the marginal tax rate is very high @ PL.
Thus there is no work incentive to rise above PL as one is actually worse off.
2. In plan (ii) we provide a subsidy to all BPL families so as to take their income to a minimum threshold.
This means that the income received doesn't change for any additional earning (as long as one
remains a BPL) and thus the marginal tax rate is 100% and there is no incentive to work.
1. In plan (iii), we vary plan (ii) such that the reduction in subsidy is not @ the same rate as increase in
income so that as one earns more his total income increases. This gives an effective marginal tax rate
of < 100% and creates an incentive to work as compared to plan (ii).
2. In plan (iv) we increase the subsidy as a person earns more up to a limit (B) and after that we reduce it
with increasing income till a new limit (B') post which it is zero. The marginal tax rate in such a plan
will be negative until B, positive till B' and 0 thereafter. It will create strongest incentive for the poor to
work but is unacceptable from equity perspective.
On Production
1. Keynes' multiplier: But it can't work @ full employment level.
2. Crowd in: If expenditure is directed towards infrastructure projects, R&D and human capital.
On Distribution
1. Social security measures, subsidies / negative income tax, social infrastructure.
2. Encouragement to labor intensive industries.
Output Gap
1. When there is a lack of AD and resources lie idle government can increase its spending on consumer
goods to raise AD. This will induce private players to utilize resources fully and raise output. But this
will not raise investment and will not lead to higher economic growth. On the other hand public debt will
rise. But a contrary view is that as AD rises private investment will increase and thus government
expenditure will crowd in I.
Domar Condition
1. He argues if r < Gy i.e. if interest rate is less than GDP growth rate then deficit financing is sustainable.
This is because GDP growth will increase income a part of which can be used to pay interest. When
this condition is satisfied a country will not fall into debt trap.
2. He argues even if economy is working @ full capacity, an increase in government expenditure will add
to income via new capacity addition as well as increased demand. Now because of this increased
income, accelerator principle will ensure private investment comes in as well. Thus both public and
private investments increase and there is higher growth in income. This can be sustainable if r < Gy.
Q. What is internal debt trap? How do the economies of developing countries fall under this trap? (2006, I, 20)
Q. "The objective of fiscal stabilization has become difficult to be achieved in most of the developing countries
due to economic compulsions and political pressures." Comment upon the statement. (2006, I, 60)